Speaker
Description
The mammary gland, originating from the ectoderm, is primarily responsible for milk production and, in humans, also has aesthetic and psychosocial importance. Breast size and contour vary widely among individuals, and complete symmetry is uncommon. Anatomically, the gland is composed of lobules that produce milk, connected by ducts that drain toward the nipple, and supported by fibrous stroma and adipose tissue.
Mammography commonly detects breast calcifications, which are categorized as benign, indeterminate, or malignant based on their morphology and distribution. Benign calcifications include skin, vascular, coarse, rod-like, round, punctate, lucent-centered, eggshell, suture, and milk of calcium types. Amorphous and coarse heterogeneous calcifications are generally considered suspicious with intermediate probability, whereas fine pleomorphic and fine linear calcifications are more frequently associated with malignancy. Diagnostic assessment also depends on distribution patterns—such as diffuse, regional, clustered, linear, or segmental—as well as size, number, and interval stability. Suspicious findings require high-quality mammographic evaluation and are often confirmed through stereotactic core needle biopsy.
Texture analysis represents a promising imaging approach because it evaluates the characteristics of the surrounding breast tissue without the need for precise calcification segmentation, thereby minimizing observer dependency. The differentiation between benign and malignant lesions relies on analyzing tissue patterns around microcalcifications and is ultimately verified by histopathological and immunohistochemical examination.
This study aims to implement texture analysis techniques on mammographic images to detect and classify breast calcifications. The objectives include distinguishing benign from malignant lesions, assessing breast tissue composition, and enhancing early detection. By incorporating artificial intelligence and computer-aided diagnosis (CAD) systems, the study seeks to improve diagnostic accuracy, assist radiologists in clinical decision-making, and contribute to more standardized and reliable breast cancer detection protocols.